Intended for healthcare professionals


Infection in xenotransplantation

BMJ 2000; 321 doi: (Published 23 September 2000) Cite this as: BMJ 2000;321:717

Studies with cell free virus are needed to define infection—there is no proof yet of safety or danger

  1. Jay A Fishman, associate professor of medicine (jfishman{at}
  1. Infectious Disease and Transplant Units, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114, USA

    Xenotransplantation is the transfer of viable cells, tissues, or organs between species. It has been proposed as a solution to the shortage of human organs (allografts) to treat people with organ failure. There are still serious immunological barriers to the broad clinical application of this technology. Concern has centered on the risk of introducing novel pathogens derived from animals into human recipients of xenografts. In addition, there is the possibility that these infections could spread from the xenograft recipient to the general population. At present, few data address the degree of risk for such interspecies infections in humans.

    Infectious diseases are common after organ transplantation, largely due to the immunosupressive agents given to prevent graft rejection.1 The risk of infection may be greater in xenograft recipients because of the possible need for greater levels of immune suppression and the role of novel pathogens from animals. 2 3 Many lists of “potential human pathogens from swine” have been created, ranging from lists of organisms likely to cause disease in transplant recipients who are immunosupressed to exhaustive lists of every organism known to infect pigs.

    Many organisms may be excluded from closed herds of donor swine by careful breeding. The porcine endogenous retroviruses are integrated into the genome, and are thus in every cell of every pig, and are therefore less susceptible to exclusion by careful breeding. There are at least three variants of porcine endogenous retroviruses (A, B, B1, and C) in native pig cells.4-6 Porcine endogenous retroviruses infect human cells in vitro and have been cloned. 4 5 Recent data suggest that despite the presence of many fragmentary copies of virus sequences, there are relatively few4-6 full length copies of the viral DNA in each cell that are capable of producing infective virus (Scobie L et al, Transplantation Society congress, Rome 2000). In addition, some genomic sites produce incomplete viral transcripts, which are not thought to be infective.5 This small number of intact genes might allow inactivation of proviruses of porcine endogenous retroviruses through genetic manipulation. Given the variability of virus production between strains of swine, it is perhaps not surprising that an inbred strain of mini-swine may produce levels of the most infectious strains of porcine endogenous retroviruses (A and B) too low to be detected by current in vitro assays (C Patience, personal communication).

    The role of receptors in infection

    Recent reports suggest that porcine endogenous retroviruses may infect human pancreatic islet cells in vitro and murine cells in vivo.7 This would suggest that both human and mouse cells carry receptors needed for infection with the virus.8 Alternatively, it is possible that the infection of murine tissues occurs only because the virus is “rescued” by one of the many known murine endogenous viruses (pseudotyping), which allows the virus to enter murine cells in the absence of a receptor for porcine endogenous retrovirus. Human cells carry these receptors but lack comparable endogenous retroviruses. Receptor studies and sequence analysis of isolates of porcine endogenous retrovirus from mice would address the basis of infection in mice. In contrast with the murine studies, in vivo studies in baboons and in humans with limited exposure to porcine cells do not show infection (Martin U et al, Transplantation Society congress, Rome 2000).9

    How then do we assess the risk of infection of porcine endogenous retrovirus in humans? There are several alternate explanations. The simplest is that human and non-human primate infections are not occurring, or the level of infection is undetectable by available assays. Or human infections may be occurring but are transient. Such transient viraemia might integrate viral DNA into the genomes of some cells of the host (patient) with the potential for “later” reactivation. Thus, porcine endogenous retroviral DNA would be found only in those cells most susceptible to infection. Assays of viral expression in peripheral blood and most tissues could be “negative” despite the presence of infection elsewhere.

    It is possible that human and primate infections are occurring but infection is masked by a gross excess of virus produced by circulating or transplanted pig cells (‘chimerism”). Present technology does not allow the detection of single human cells secreting virus. Thus, in a population of cells including both human and pig cells, infection with porcine endogenous retroviruses is detected only when the amount of virus produced exceeds that amount expected from a normal pig cell. Such assays rely on ratios of porcine endogenous retroviral DNA or mRNA to normal pig gene products to suggest that porcine endogenous retrovirus is produced “in proportion” to the number of pig cells. How much virus is expected? What is normal for each type of pig cell transplanted into a primate? Few data exist to guide such interpretation.

    In in vivo infections we rely on indirect measures to determine that murine cells are infected by porcine endogenous retroviruses and that humans and baboons are not so infected. Are these observations contradictory? Not really. Technology imposes limits on our ability to detect virus. However, the amount of porcine endogenous retroviral mRNA produced appears to vary with the strain of pig, the tissues tested, and possibly with environmental factors (for example, cytokines, immune responses).

    It is critical that the presence or absence of infection of humans or baboons be defined using purified virus in the absence of pig cells. Otherwise, data may be overinterpreted beyond the power of the assays being performed. These data are then reinterpreted as proof of the relative safety or as proof of the danger of porcine xenotransplantation. As yet, such final proof, either way, does not exist. Rapid progress is being made in defining the biology of porcine endogenous retroviruses. Further studies are still needed to define the infectious risks associated with xenotransplantation into humans.


    JAF is a consultant to Biotransplant, Novartis, Genzyme, Fujisawa, and Gilead Incorporated.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    View Abstract